Message from discussion
DataFlow - Release Update
Received: by 10.180.83.166 with SMTP id r6mr2643465wiy.1.1335284557074;
Tue, 24 Apr 2012 09:22:37 -0700 (PDT)
X-BeenThere: dataflow-devel@googlegroups.com
Received: by 10.216.199.199 with SMTP id x49ls9757474wen.1.gmail; Tue, 24 Apr
2012 09:22:34 -0700 (PDT)
Received: by 10.216.137.196 with SMTP id y46mr1365140wei.7.1335284554958;
Tue, 24 Apr 2012 09:22:34 -0700 (PDT)
Received: by 10.216.137.196 with SMTP id y46mr1365139wei.7.1335284554934;
Tue, 24 Apr 2012 09:22:34 -0700 (PDT)
Return-Path: <stepha...@criticalec.com>
Received: from mail-wg0-f45.google.com (mail-wg0-f45.google.com [74.125.82.45])
by gmr-mx.google.com with ESMTPS id ga1si5056590wib.2.2012.04.24.09.22.34
(version=TLSv1/SSLv3 cipher=OTHER);
Tue, 24 Apr 2012 09:22:34 -0700 (PDT)
Received-SPF: neutral (google.com: 74.125.82.45 is neither permitted nor denied by best guess record for domain of stepha...@criticalec.com) client-ip=74.125.82.45;
Authentication-Results: gmr-mx.google.com; spf=neutral (google.com: 74.125.82.45 is neither permitted nor denied by best guess record for domain of stepha...@criticalec.com) smtp.mail=stepha...@criticalec.com
Received: by wgbdt14 with SMTP id dt14so669603wgb.2
for <dataflow-devel@googlegroups.com>; Tue, 24 Apr 2012 09:22:34 -0700 (PDT)
d=google.com; s=20120113;
h=mime-version:x-originating-ip:date:message-id:subject:from:to
:content-type:x-gm-message-state;
bh=g2K8YsBgOq+1PN0ci8G4phGAdkDlnNW5/OHigktJqKU=;
b=JNxKs3szy9RqVV89qKsypFMvDicmXLZY+bXg69jsij4xsFdmNyX+4T7aTx19JeiU9I
oi/odPUm4j9NMO/kTs2S7vwC3J8jb3C+SgYElLXMuF9RsvasYjQjPDaiCtNw0rIWLAql
TFtCA/FdAypCsn6fsF7CkqaJ6bdHk9+5hptW+5HG0Qnv6i56Myiq9mYVK7eW18PjJBnr
2b/THPSfP3QrpwbToya2P1j9Pdv1k0yas+gOSVQi+tyUvkc53AYbUDY7BLP+II6bSCRc
M1fIOzIsQNSLqcePNJ5lPpSbMtqZGbCssrQc2jPz5SuCEZdH8vWCLNzUVHkYIxfOc9op
5B2w==
MIME-Version: 1.0
Received: by 10.180.90.233 with SMTP id bz9mr1672316wib.21.1335284554541; Tue,
24 Apr 2012 09:22:34 -0700 (PDT)
Received: by 10.180.93.196 with HTTP; Tue, 24 Apr 2012 09:22:34 -0700 (PDT)
X-Originating-IP: [89.242.250.236]
Date: Tue, 24 Apr 2012 17:22:34 +0100
Message-ID: <CALftN-u37NcezBNY_uGhx7u6zMybpj=fgo2_A4GvCVMJ-yq...@mail.gmail.com>
Subject: DataFlow - Release Update
From: Stephanie Taylor <stepha...@criticalec.com>
To: dataflow-devel <dataflow-devel@googlegroups.com>
Content-Type: multipart/alternative; boundary=f46d043be2728c2c1804be6f294b
X-Gm-Message-State: ALoCoQk3X2cBxAFq7bVZ4tsi3W0rvcBDKW6DC1XuAiWkUF2gTHhXVHPUh1+heZTMuYJnWkDgFQR3
--f46d043be2728c2c1804be6f294b
Content-Type: text/plain; charset=ISO-8859-1
Hi,
Further to my previous email, Richard is working with Anusha and Bhavana to
resolve the current issues.
We'll update everyone tomorrow afternoon.
Here is more info from Richard on the specific issues -
Hi Folks,
The error is quite peculiar, we are having some trouble pinning it
down and replicating. Status at the moment
1/ I found a minor bug in the client which - if it isn't causing
problems now - would cause problems in the future. This is now fixed,
and committed to the new github repo for the python client:
https://github.com/swordapp/python-client-sword2
2/ We are finding intermittent and not easily reproducible XML parsing
errors in databank. This is causing databank to return a 500 to
deposit requests, which is likely the real cause of the problem
3/ DataStage is not attempting to auto-discover service documents;
this has no bearing on the deposit error, but it's unclear to me just
yet why this is. Service document URLs must - for the time being - be
entered manually into the repository configuration
4/ DataStage does not have a mechanism for selecting deposit
collection, which is the reason we are not seeing deposits come
through to the sandbox - they are, in fact, going through to the first
collection in the databank service document, which is "demo"; see:
http://databank-vm1.oerc.ox.ac.uk/demo
So, where next? I have to try and figure out why databank is giving
us 500 errors, which I can only do with a current databank install.
Since I'm still running the swordification branch, I need to put a bit
of time into upgrading; hopefully this will be quick and simple, but
if there are large changes that I need to take into account this might
take a bit longer. Once i have that up, I have to try to reproduce
the error and then debug it. Assuming it is repeatable, I hope that
I'll be able to get it fixed relatively rapidly.
Meanwhile, I also want to try to connect DataStage to my testing sword
server, to make sure that there are no other issues at that end.
I'd anticipate that this work will take at least all of tomorrow; I'll
update you tomorrow afternoon on state of play and try to estimate
when I can get it fixed up.
Cheers,
Richard
--
Stephanie Taylor
Director
Critical Eye Communications Limited
Registered Office:
113 Dye House Lane
New Mills
High Peak
SK22 4BA
UK
M: +44 (0)7745 123705
E: stepha...@criticalec.com
W: http://www.criticalec.com
Registered as a Company in England & Wales.
Company Number: 5871413
--f46d043be2728c2c1804be6f294b
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Hi,<div><br></div><div>Further to my previous email, Richard is working wit=
h Anusha and Bhavana to resolve the current issues.</div><div>We'll upd=
ate everyone tomorrow afternoon.</div><div><br></div><div>Here is more info=
from Richard on the specific issues -=A0</div>
<div><br></div><div><span style>Hi Folks,</span></div><br style><span style=
>The error is quite peculiar, we are having some trouble pinning it</span><=
br style><span style>down and replicating. =A0Status at the moment</span><b=
r style>
<br style><span style>1/ I found a minor bug in the client which - if it is=
n't causing</span><br style><span style>problems now - would cause prob=
lems in the future. =A0This is now fixed,</span><br style><span style>and c=
ommitted to the new github repo for the python client:</span><br style>
<br style><a href=3D"https://github.com/swordapp/python-client-sword2" targ=
et=3D"_blank" style>https://github.com/swordapp/python-client-sword2</a><br=
style><br style><span style>2/ We are finding intermittent and not easily =
reproducible XML parsing</span><br style>
<span style>errors in databank. =A0This is causing databank to return a 500=
to</span><br style><span style>deposit requests, which is likely the real =
cause of the problem</span><br style><br style><span style>3/ DataStage is =
not attempting to auto-discover service documents;</span><br style>
<span style>this has no bearing on the deposit error, but it's unclear =
to me just</span><br style><span style>yet why this is. =A0Service document=
URLs must - for the time being - be</span><br style><span style>entered ma=
nually into the repository configuration</span><br style>
<br style><span style>4/ DataStage does not have a mechanism for selecting =
deposit</span><br style><span style>collection, which is the reason we are =
not seeing deposits come</span><br style><span style>through to the sandbox=
- they are, in fact, going through to the first</span><br style>
<span style>collection in the databank service document, which is "dem=
o"; see:</span><br style><br style><a href=3D"http://databank-vm1.oerc=
.ox.ac.uk/demo" target=3D"_blank" style>http://databank-vm1.oerc.ox.ac.uk/d=
emo</a><br style>
<br style><br style><span style>So, where next? =A0I have to try and figure=
out why databank is giving</span><br style><span style>us 500 errors, whic=
h I can only do with a current databank install.</span><br style><span styl=
e>Since I'm still running the swordification branch, I need to put a bi=
t</span><br style>
<span style>of time into upgrading; hopefully this will be quick and simple=
, but</span><br style><span style>if there are large changes that I need to=
take into account this might</span><br style><span style>take a bit longer=
. =A0Once i have that up, I have to try to reproduce</span><br style>
<span style>the error and then debug it. =A0Assuming it is repeatable, I ho=
pe that</span><br style><span style>I'll be able to get it fixed relati=
vely rapidly.</span><br style><br style><span style>Meanwhile, I also want =
to try to connect DataStage to my testing sword</span><br style>
<span style>server, to make sure that there are no other issues at that end=
.</span><br style><br style><span style>I'd anticipate that this work w=
ill take at least all of tomorrow; I'll</span><br style><span style>upd=
ate you tomorrow afternoon on state of play and try to estimate</span><br s=
tyle>
<span style>when I can get it fixed up.</span><br style><br style><span sty=
le>Cheers,</span><br style><br style><span style>Richard</span><div><div><b=
r></div>-- <br>Stephanie Taylor<br>Director<br>Critical Eye Communications =
Limited<br>
Registered Office:<br>113 Dye House Lane<br>New Mills<br>High Peak<br>SK22 =
4BA<br>UK<br>=A0<br>M: +44 (0)7745 123705<br>E: <a href=3D"mailto:stephanie=
@criticalec.com" target=3D"_blank">stepha...@criticalec.com</a><br>=A0<br>W=
: <a href=3D"http://www.criticalec.com" target=3D"_blank">http://www.critic=
alec.com</a><br>
=A0<br>Registered as a Company in England & Wales.<br>Company Number: 5=
871413<br><br>
</div>
--f46d043be2728c2c1804be6f294b--